The invention relates to xe2x80x9crakexe2x80x9d receivers which have the ability to combine received signals having different transmission delays, and more particularly to rake receivers capable of predicting short-term changes in received signal strength from various transmission paths and dynamically adapting thereto.
Code-division multiple access (CDMA) radio systems are well known.
FIG. 2 depicts portions of a conventional receiver that would typically be used to receive and demodulate the CDMA signal produced by the conventional transmitter of FIG. 1.
The receiver must generate the same long code (LC) and pseudo-noise (PN) sequences as the transmitter for demodulating the received signal. However, transmissions are generally affected by transmission delays. Thus, the receiver must determine the delay in order to determine a starting position within the LC and PN sequences for demodulating the received signal. This determination is generally performed by a xe2x80x9csearcherxe2x80x9d. If multipath components are being received with sufficient signal strength, a typical searcher can determine the intercomponent delays.
A conventional xe2x80x9crake receiverxe2x80x9d, adapted from the receiver of FIG. 2, provides replicated circuity for each of the multipath components. Such a receiver is shown in FIG. 3. Where the receiver of FIG. 2 has a single digital correlator 2 (depicted in the lower half of FIG. 2), FIG. 3 includes multiple digital correlators 2, called xe2x80x9cfingersxe2x80x9d and numbered 2-1, 2-2, . . . 2-n. The number of fingers in a rake receiver is typically on the order of two to five.
Each finger has a delay circuit (3-1, 3-2, . . . 3-n) preceding it, and each finger""s output is input to a summer 6 through a weighting circuit (5-1, 5-2, . . . 5-n).
Searcher 4 determines the transmission delay from the receiver to the transmitter. In the case of multipath transmission, it inherently determines the delay and signal strength for each of the multipath components. Information pertaining to path delay and power (collectively known as the Channel Impulse Response (CIR)) can be used to set the initial delays 3-1 etc. associated with each finger. FIG. 9 depicts an exemplary received signal comprising two multipath components 501 and 502. Component 502 arrives later (delayed by an amount connoted xe2x80x9cdeltaxe2x80x9d) and with a different signal strength. Both components exhibit Rayleigh fading, well known in the art.
Searcher 4 can determine the delay xe2x80x9cdeltaxe2x80x9d amounts associated with each multipath component and accordingly provide information for configuring the delays in FIG. 3 (expressed in terms of a number of PN chip durations). In this instance, delay 3-1 would be set to zero (to handle component 502) and delay 3-2 would be set to the amount xe2x80x9cdeltaxe2x80x9d so that component 501, estimated to arrive earlier by that amount, will be delayed better to coincide with component 502.
Finger assignment module 9 receives the raw channel impulse response (CIR) estimate from searcher 4 and from it determines how best to place the fingers, i.e., what delay the fingers should be set to in order to optimize receiver Frame Error Rate performance. Finger assignment module 9 is apprised of the current state of the fingers, and assigns and deassigns fingers according to the current signal conditions as determined by the searcher in conjunction with the current state of the fingers.
FIG. 3 also shows that each finger has provision for setting the weight input to its weighting circuit 5-n. It does so according to an estimate of the Eb/No (energy per bit divided by noise, a signal-to-noise ratio) of the multipath component assigned to the finger. Conventional calculations may be used for calculating the Eb/No estimate, such as the ratio of the peak correlation value to the xe2x80x9cnoise floorxe2x80x9d in the Fast Hadamard Transformsxe2x80x94see elements 209 and 210 of FIG. 2.
FIG. 4 shows the internals of the conventional finger assignment module 9. The raw CIR estimate from the searcher is received in block 901 and put through a Wiener filter. The results for each multipath component are input to a time-averaging filter (of which 902A is exemplary) in filter bank 902. Block 903 determines a power delay profile, which is a more refined estimate of the CIR in that it is a time average whereas the CIR reflects an instantaneous state.
The power delay profile produced in block 903 is passed to assignment computation algorithm 905. Block 904 keeps track of the current state of the fingers. Both are connected to finger assignment command generator 906, which generates commands issued to the fingers of the rake receiver.
Under the example of FIG. 9, two fingers would be assigned and the rest deassigned. When summer 6 combines the weighted output of finger 2-1 and the delayed, weighted output of Finger 2-2 the result will be more reliable than if just one of the multipath components were used, provided that the delays and weights are set correctly.
The conventional ways of setting the weights, based on current signal strength, do not account for short-term variations in signal strength. An improvement in performance could be attained by providing weights more dynamically responsive to short-term variations.
Accordingly, there exists a need for a receiver that can dynamically predict quality of multipath components over short time periods in a highly time-variant environment.
It is thus an object of the present invention to provide an improved CDMA receiver.
It is another object of the present invention to provide a CDMA receiver with improved error-rate performance for high-data-rate systems operating within highly time-variant environments.
It is yet another object of the present invention to provide a CDMA receiver with the ability to predict short-term variations in strength of multipath components and to compensate for those variations.
These and other objects of the invention will become apparent to those skilled in the art from the following description thereof.
These and other objects may be accomplished in a CDMA receiving apparatus having a rake receiver with several fingers, each for demodulating a multipath component of a CDMA signal, and having a searcher and finger assignment means for assigning multipath components to fingers and for initially setting finger parameters, by the present systems and methods of predicting short-term variations in strength of each multipath component, and using those predictions to assign and deassign fingers to multipath components and to initialize delays associated with those multipath components and for determining weighting of each multipath component.
An embodiment of the invention provides a rake receiver and finger assignment means for assigning multipath components to the fingers of the rake receiver. The finger assignment means includes predictors for determining power delay profile in a predictive manner from a channel impulse response determination, and assigning multipath components to fingers of the rake receiver and setting initial finger delay accordingly. In the rake receiver, a predictor is associated with each of the fingers. The predictor receives estimates of signal-to-noise ratio from its associated finger and also receives estimates of the channel fade rate, and estimates short-term fade rate of a multipath component from those data. The weight of each multipath component is set according to those estimates .
The invention will next be described in connection with certain exemplary embodiments; however, it should be clear to those skilled in the art that various modifications, additions and subtractions can be made without departing from the spirit or scope of the claims.